root-cause-analysis

Decision Router

IF problem has a clear single causal chain from symptom to root → Five Whys (drill down iteratively, stop at systemic cause) IF problem has multiple potential contributing factors across domains → Fishbone / Cause-and-Effect (explore 6 categories systematically) IF "human error" appears as a cause → keep digging regardless of method — the system should have made the error impossible IF combining with structured documentation → produce findings as input for section 4 of an A3 analysis

Root Cause Analysis

Root cause analysis methods for debugging incidents, process failures, and quality problems. Choose the method that matches the problem structure.

Core Principle

Every surface symptom is the end of a causal chain. Treating the symptom leaves the chain intact. The goal is to find the systemic cause — a missing validation, unclear process, absent automation, or design choice.

Five Whys

Iteratively ask "why" to trace from surface symptoms to fundamental systemic causes. The name "five" is a guideline — stop when you reach a process, system, or policy cause.

Process

1. State the problem clearly — the specific symptom
2. Ask "Why did this happen?" — answer with the direct cause
3. For each answer, ask "Why?" again — trace one level deeper
4. Repeat until reaching a systemic cause (missing process, validation, automation, or design gap)
5. Validate by tracing forward — root cause → symptom should be a logical chain
6. If branches emerge, explore each independently
7. Propose solutions addressing the root cause — not intermediate answers

Output

A causal chain from symptom to root cause with solutions tied to the root cause level.

Fishbone / Cause-and-Effect

Systematic exploration of all potential causes across six categories using the Ishikawa method.

The Six Categories

Category	What to Examine
People	Skills, training gaps, communication, team dynamics
Process	Workflows, procedures, standards, reviews
Technology	Tools, infrastructure, dependencies, config
Environment	Workspace, deployment targets, external factors
Methods	Approaches, patterns, architectures, practices
Materials	Data quality, third-party services, inputs

Process

1. State the problem clearly — the "head" of the fish
2. For each category, brainstorm potential causes (don't stop at the first one)
3. Identify which causes are symptoms vs. root causes
4. Cross-reference — mark causes that span multiple categories (systemic)
5. Prioritize — score by impact and likelihood
6. Propose solutions tied to specific root causes

Output

A categorized cause map identifying root causes across six dimensions with prioritized solutions.